Dynamic Programming Algorithm for Nucleic Acid Sequence Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for aligning nucleic acid sequences struggle with handling overlapping reading frames and frameshifts, leading to unrealistically large numbers of artificial frameshifts and loss of biological meaning in amino acid sequences, especially when dealing with sequences containing coding, non-coding, and possibly overlapping reading frames.
Innovation Solution
A dynamic programming algorithm that locally optimizes a partial alignment score based on nucleotide substitution costs, amino acid substitution costs, gap insertion costs, and frameshift penalties, allowing for simultaneous alignment of multiple coding and non-coding sequences with a single recursive build-up of a two-dimensional array, effectively handling overlapping reading frames and accounting for occasional frameshifts in O(mn) computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing alignment methods are used on sequences with overlapping reading frames, then alignment is performed, but unrealistically large numbers of artificial frameshifts occur and biological meaning is lost
Solution Approach 1:
The patent changes the parameter of frameshift penalty from a fixed value to a dynamic value that varies based on reading frame context. By implementing state-dependent penalty parameters that recognize coding versus non-coding regions and overlapping frames, the algorithm adapts its scoring criteria to preserve biological meaning while maintaining alignment accuracy.
Solution Approach 2:
The patent introduces an intermediary layer of reading frame annotation and state tracking between the raw sequence alignment and the final scoring. This intermediary system identifies coding regions, non-coding regions, and overlapping frames, then uses this information to modulate alignment penalties, thereby preventing artificial frameshifts in coding regions while allowing flexibility in non-coding regions.
2Reliability
If multiple sequence alignment methods are used, then comprehensive alignment is achieved, but computational complexity increases significantly
Solution Approach 1:
The patent segments the alignment problem into independent dynamic programming subproblems for each reading frame, then combines the results. By dividing the complex multiple sequence alignment into frame-specific alignments and using efficient DP recurrence relations, the method achieves comprehensive alignment coverage while maintaining O(mn) computational complexity for pairwise alignments.
Solution Approach 2:
The patent performs partial alignments for each reading frame separately rather than attempting a single comprehensive alignment of all frames simultaneously. This partial action approach allows the algorithm to process each frame with appropriate penalties while avoiding the exponential complexity of simultaneous multi-frame alignment, then integrates results to achieve comprehensive coverage.
3Measurement precision
If frameshift penalties are applied to all gaps, then alignment precision is improved, but overlapping reading frames cannot be handled correctly
Solution Approach 1:
The patent applies local quality by making frameshift penalties context-dependent rather than uniform. Coding regions receive high frameshift penalties to maintain precision, while non-coding regions and overlapping frame regions receive reduced or zero penalties to allow flexibility. This local differentiation of penalty quality enables the algorithm to maintain precision where needed while adapting to the structural complexity of overlapping frames.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The current invention concerns a computer-implemented method, a computer system and a tangible non-transitory computer-readable data carrier comprising a computer program product for aligning a query nucleic acid sequence against a reference nucleic acid sequence. An optimal alignment score of the query sequence against the reference sequence is determined by a dynamic programming algorithm. The algorithm is configured for locally optimizing a partial alignment score based on a nucleotide substitution cost, a gap insertion cost, an amino acid substitution cost, as well as a frameshift penalty.